[Hemorrhagic gastro-duodenal ulcers. In which cases should emergency surgery be performed?]. / Ulcères gastro-duodénaux hémorragiques. Lesquels faut-il opérer en urgence?

From January 1983 to December 1987, 127 patients with bleeding peptic ulcer were admitted to hospital. The mean age of the 85 males was 57 years and 72 years for 42 females. All but four of the patients were managed medically after emergency endoscopy. Twenty-seven patients required surgical operations (21.2%): seven for cataclysmic haemorrhage, eight for persistent haemorrhage, twelve for recurrent bleeding. An analysis of factors leading to the necessity of surgical haemostasis was undertaken by considering the clinical status, endoscopic findings and laboratory results. The size of the ulcer (greater than 2 cm) was the most significant parameter (less than 0.01). Five other criteria (rectal bleeding) shock, endoscopic signs of recent haemorrhage, gastric or duodenal posterior ulcer) were also significant (p less than 0.05). Considering the gravity of these patients (six deaths among twenty-seven), clinical trials in bleeding peptic ulcer disease should only include patients in the high risk group.

[Pancreatic fistula following cephalic duodenopancreatectomy. Directed fistulization or total pancreatic exeresis?]. / Fistules pancréatiques après duodénopancréatectomie céphalique. Fistulisation dirigée ou totalisation de l'exérèse pancréatique?

A frequent complication (8.5 to 52.8%) of pancreatoduodenectomy (PD) for cancer, pancreatic fistula (PF) is difficult to treat, and an analysis of 30 cases of PD (27 for cancer, 3 for chronic pancreatitis) is used to determine risk factors and most effective therapy. Fistula developed in 6 cases (20%) and three risk factors were determined: preoperative renal impairment and hypoalbuminemia and ligature of pancreatic stump. Although not statistically significant, other factors--cancer, emergency surgery, fragile pancreatic tissue, thin Wirsung, pancreatojejunal anastomosis, absence of decompression of the raised jejunal loop--in this small series nevertheless provoked a marked increase in PF. One patient recovered after medical treatment, all five patients operated upon by whatever technique failing to survive. This agrees with literature data indicating heavy mortality (44.4 to 100%) after surgery. This should therefore be reserved for cases of PF failing to respond to adequate medical treatment, or with hemorrhage or intra-abdominal sepsis not controlled medically. The most effective therapy for PF is prophylactic, combining selection of patients as a function of risk factors, and treatment of pancreatic stump adapted to caliber of Wirsung and quality of remaining pancreatic tissue.